openSUSE Strategy Discussion: openSUSE - #1 KDE distribution

Hi all!

The next strategy that came from the community is “#1 KDE distribution”.
Here you are!

—8<------8<------8<------8<------8<------8<—

=== openSUSE - #1 KDE distribution ===

== Statement ==

There is a powerful movement towards simple and easy software targeting
end users. Most distributions focussing there use Gnome for it, but KDE
appplications also have a lot of potential. The technology is there,
it’s just, by default, cluttered and complicated in some areas. openSUSE
has always had a great KDE based distribution, it’s our stronghold. Like
most distributions, we shipped a mostly vanilla desktop. We polished
some area’s, stabilized and backported some upstream patches but not
much more. Only Mandriva currently ships a heavily modified KDE based
desktop, making it look and work like the previous 3.x series. There is
surely room for a distribution which changes the focus of KDE software
more to end users, making choices upstream finds hard to implement. This
means mostly changing default configuration, choices of applications
etc, all in all setting up an environment end users should be far more
comfortable in.

The proposal is thus: let’s focus the openSUSE efforts on KDE as the
main application and desktop provider and put a lot of effort in
customizing, simplifying and polishing it to be ready for end users. Of
course we should also make sure we cater as good as possible for Qt and
KDE developers, giving for example MeeGo developers a welcoming home.

Note: This proposal is not meant to start another flamewar, nor to drop
GNOME or any other desktop environment (of course there still will be
GNOME, LXDE and Xfce Spins available), it’s just about our primary focus.

== Activities ==

= We need to be excellent in the following =

  • create a visual unique desktop experience with KDE
  • adopt the latest KDE technology early, but care about stability
  • offer the best development platform for Qt/KDE developers
  • work together with the KDE community to establish a compatible release
    cycle
  • integrate MeeGo (focus on Qt)
  • offer a LiveCD spin which contains the latest KDE version (like the
    KDE Four Live) but a bit more official)
  • lobby for KDE
  • offer KDE reference implementation(s) and the tools to build them
    (OBS, Suse Studio)
  • improve KDE upstream
  • create a more usable experience by tuning defaults and having a better
    selection of software
  • market our product as a great, easy end user distribution
  • work to integrate technologies like ownCloud, OCS, Social Desktop and
    a GHNS centric “AppStore” in openSUSE

= We will try to do the following effectively =

  • deliver a build service for building distribution and applications
  • provide best desktop experiences with KDE and MeeGo (focus on Qt)
  • offer an good platform for Java, Ruby, Python and other developers
  • improve Kolab integration to offer a complete groupware suite (client
    and server)
  • create Qt user interfaces for apps which only offer a GTK frontend
  • bugfixing
  • testing
  • collaborate with other Linux distros
  • support community-led spins of other desktop environments to benefit
    from the rich spectrum of Linux software available
  • improve integration of non-KDE applications (GTK, wxwidgets, Wine)
    with KDE file dialogs, theming, etc

= As project, we will not focus on the following anymore =

  • trying to be the best in ‘everything’, but in tasks where we already rule
  • server stuff?

I like the idea!

Btw. as strategy team we had a hard time with this especially due to its divisiveness and for this reason hesitated to put it for public discussion but did not want to censor it either.

Clear guidance is important. When a newbie asks whether Gnome or KDE is better, we say that they are equivalent and that it is a personal preference. However, when they hear that there are a Distributions that clearly prefer Gnome, while there is none that prefers KDE, their decision is made.

People want help with such choices and a clear stance is helpful to everyone involved.

(Sorry I just posted something similar in the KDE/Gnome poll thread, but I think here it is more appropriate.)*

You know that openSUSE 11.2 during installation sets the default to KDE so that people do not need to make a decision. Still, both are Tier 1 desktops and both are great.

This strategy has some nice advantages but also brings a significant risk.
Find my visualization of the strategy here: http://mevin.net/download/options/options-No1%20KDE%20distribution.pdf

Advantages:

  • Focus: This is one of the strategies that really focuses on particular segments and helps to save resources.
  • Differentiate: In contrast to the many GNOME favoring distros this could be a USP and a good way to differentiate from other distros.
  • Popularity of Qt and KDE is taken into account.

Risks: The main risk is what we loose, when we go for that strategy.

  • Aren’t we moving to far away from the core market and the real strength of Linux, the server market? To chase away the cash cow might have consequences.
  • How many of the other DE developers, users, contributors, fanboys and -girls do we loose and how many are joining instead? Wouldn’t it be better to keep both? Isn’t the fragmentation of the FLOSS community one of the big threads? Yes, it is (http://www.mevin.net/download/openSUSE-strategy_02.pdf)

Summary:
The strategy clearly differentiates and concentrates. The risk are very high. The strategy is working against one of the main challenges (collaboration within FLOSS).

What could we do?
Minimize the risks:
We should be an open community, welcoming KDE as well as GNOME, LXDE, … underlining from time to time the KDE friendliness (compared to GNOME focusing distros).

User the strength:
I think it could be very helpful to transform this ideas into a sub-strategy for the openSUSE KDE team. There are really good points in it and it gives a good overview what is possible and points to a direction for the sub unit.

+1
I think it is essential for a community to discuss such points or they will never stand wholeheartetly behin the new strategy. Uncleared issues will always come back and hamper the forthcoming of a community.

= As project, we will not focus on the following anymore =

  • trying to be the best in ‘everything’, but in tasks where we already rule
  • server stuff?

I personnaly find these quiet ridiculous. If we don’t have any effort in server, how would you incoporate Kolab, and offer a THE one package many smb are looking for.
Are you really ready to leave the world open to ubuntu, osx, and windows ?

openSUSE has a long history of stability particularly in server usage, don’t break what’s is working.
Now in the desktop part, especially the kde one (which is mine from a long time, and will be surely in the future) what it’s needed is somewhere things that work.
Remember one of those AGiLE rules : commit only on no error when all tests works. Helping upstream to write tests, and make them runable, will benefit to everybody. Dev’s contributors, and finally to users.
You talk about the “social” things, but sorry I’m waiting from the last 2 years a semantic desktop like it was promise 2.5 years ago ! Finish the work, stabilize, make a new step, finish the work, stabilize, …

I offer myself to work on that, im not developer, im an architect… but im been always interest on helping linux to get “design thinking” concepts, because right now design in linux is just a tool to make it look “beauty”… design has to help create a more usable experience…

Sorry, but it’s not clear to me on what you want to work? I expect “to work on that” refers to a post above but I can’t figure out which one.

A Focus on KDE “part of openSUSE” is a good thing. IMHO it does not deter the GNome, LXDE, XFCE , Server etc. People need to start thinking more than one dimensional! It took years to educate people that in a multi-tasking environment it is ok to open several apps and even several desktops as the need rises. Linux has always been a good platform for handling many OS’s as either multi-boot or virtualized yet even learned profs tend to try and dedicate one machine per OS then try and daisy chain them together to share results. To put it another way, If you spend time to improve your backyard does it really mean that you have devalued home garage or car?

  • create a visual unique desktop experience with KDE
  • adopt the latest KDE technology early, but care about stability
  • offer the best development platform for Qt/KDE developers
  • work together with the KDE community to establish a compatible release
    cycle
  • integrate MeeGo (focus on Qt)
  • offer a LiveCD spin which contains the latest KDE version (like the
    KDE Four Live) but a bit more official)
  • lobby for KDE
  • offer KDE reference implementation(s) and the tools to build them
    (OBS, Suse Studio)
  • improve KDE upstream

At last KDE will get some respect!
As someone that both helped & stuck it out with KDE4.0 I say,“YEAH!”

From a straightforward distro/product positioning viewpoint, this strategy is probably the most exciting (to me that is, as user of KDE mainly but also Gnome). Doesn’t linux also need the very “best of breed” KDE distro, to attract and keep those windows users who seek a similar but alternative approach, especially for their home systems? Most of those only hear about Ubuntu, they are attempting to integrate new technologies, so openSUSE could provide a KDE alternative by integrating new technologies.

Yes indeed, but how divisive? It begs the question: Will the implementation of this strategy leave the openSUSE project organization fit and able to continue the work required to produce the best KDE, plus reasonably good Gnome, LXDE, etc.? I am assuming that means good enough to motivate and retain appropriate Gnome/LXDE developers, and also existing users. It also means good enough for Gnome users thinking about joining a distro that seriously offers multiple DE’s, and attractive services. For example, it could be a good place for those users who believe they left a prematurely released KDE4, for a more stable Gnome experience - they may want to return to KDE now.

For this strategy to work, I also have to assume that the project will provide the alternative DE developers with the same infrastruture and services enjoyed by KDE developers. The same goes for the larger openSUSE community in recognising and welcoming the contribution of each DE to the project’s overall success.

Would this strategy of increased focus be more likely to attract more contributions from KDE developers? Can anyone reduce my uncertainty on this one.

This is a very high risk strategy unless the uncertainties (e.g. my assumptions above) can be addressed or removed, and detailed plans put in place to manage the risks.

If the anticipated risk of division can be mitigated, this strategy would be my first choice as it offers openSUSE an attractive USP.

From my previous post, you probably gather that I agreed with much of your posting here. However, not sure I fully understood the above point.

I understand that Linux enables commercial organizations such as Novell to develop, package and sell server-based system software and services. It’s therefore the “cash cow” of several companies. I don’t see openSUSE as one of those companies. I suppose it depends on the strength of the link between Novell’s server-based products and openSUSE. What is the nature of that link? How far are you stretching the elastic to arrive at “moving too far away from”? Whose “core market”?

So, openSUSE would focus on being the largest KDE contributor and advocate also providing a showcase for the latest-and-greatest KDE has to offer?

That doesn’t sound too bad. Fedora and Ubuntu just had a tiff over their respective Gnome contributions (16% vs. 1%) and KDE patiently chuggs along.

The down-side of this is that openSUSE is also the most direct method of getting access to Novell (and Mono) improvements, unless that is to be conceded to Ubuntu who seems to be the #2 supporter for Mono.

I can’t help and wonder, though, if going more KDE-centric would reflect negatively on the other openSUSE offerings including Yast? Woudlnt this basically be saying “if you want the advantage of Yast (as well as stability/capability/etc. of openSUSE) you have to move to KDE”?

I think openSUSE’s focus on the KDE offering would be good, as openSUSE has a strong reputation for “if you want KDE, go with openSUSE or Mandriva”. KDE apps have a lot to offer and is often overlooked by users. I’m not sure if it is enough in and of itself, but it is a very plausible secondary focus.

I don’t think moving away from servers is the best idea, though.

Recently somebody has pointed me to a server distro that uses the eBox web interface to configure everything. OpenSUSE with a web-based Yast could probably overcome this distro and make some headway.

I don’t think the server offering is a drag, I think it hasn’t been marketed/focused to its potential.